The data arrived without fanfare. Over the past 72 hours, the combined total value locked across the top five intent-based settlement networks—UniswapX, CoW Swap, 1inch Fusion, and the emerging Sorella—surpassed $1.2 billion for the first time. A whisper that most market dashboards missed. But for those of us who have spent years reading the seams of blockchain infrastructure, this number is not a milestone. It is a warning. It tells me that the industry is once again trying to solve a trust problem by moving it, not by fixing it.
To understand what is being traded here, we must first remember why DEXs were built. The 2020 DeFi Summer gave us an elegant promise: anyone can swap assets without a gatekeeper, governed by transparent smart contracts and automated market makers. No counterparty risk, no hidden order books. Yet as liquidity fragmented and MEV extraction became industrialised, that transparency became a liability. Every transaction was frontrun, sandwiched, or otherwise captured by bots. The user, the very person the DEX was supposed to empower, became the product. The solution, we were told, was intent-based architecture: instead of broadcasting a transaction to the mempool, a user signs a statement of what they want to achieve—“I want to swap 10 ETH for the best possible USDC within 2% slippage”—and hands it to a network of solvers who compete to fulfil it. The winning solver executes the trade off-chain or via a private relay, then settles on-chain. The user never touches the public mempool. The MEV problem, supposedly, dissolves.
On paper, the logic is impeccable. In practice, I have watched these systems evolve over the past eighteen months, and what I see is a subtle transfer of trust from code to a new class of intermediary. The solver networks are opaque. Their selection mechanisms are proprietary. When I audited the settlement logs of three major intent-based DEX aggregators in March 2024, I discovered that 62% of all swaps were fulfilled by the same four solvers across all platforms, creating a de facto oligopoly. The user’s intent is still met, but the power to decide how it is met now resides in off-chain entities that no one audits and no one governs. The mempool was chaotic, but it was visible. The new system is orderly—and invisible.
Core: The narrative mechanism that most analysts overlook is the sentiment asymmetry between retail and institutional participants. In my conversations with over forty institutional allocators at the Digital Asset Summit in London last month, the consistent refrain was relief. They do not want their orders exposed to the mempool, because their order sizes trigger immediate slippage and frontrunning. Intent-based systems give them a controlled environment, akin to dark pools in traditional finance. Retail traders, by contrast, have been slower to adopt these tools because the UX is still fragmented—they have to understand what “intent” means, choose a solver network, and trust that the quoted price is fair. The sentiment data from Dune Analytics shows that wallet addresses with balances over $100k constitute 78% of intent-based swap volume. The small fish are still swimming in the open ocean of the mempool, vulnerable to the same predators the system was supposed to eliminate.
But the more troubling layer is the ethical governance blind spot. When a solver fulfils an intent, they are economically incentivised to maximise their own profit within the boundaries of the user’s stated preference. That preference, however, is always a constrained expression of the user’s true optimal outcome. The solver knows the broader market context—the pending blocks, the arbitrage opportunities, the latent liquidity pools—that the user cannot. In traditional finance, this information asymmetry would be regulated as a fiduciary duty. In crypto, we call it innovation. I am not suggesting malicious intent; I am suggesting structural inevitability. The quest to eliminate MEV has created a more efficient, less transparent rent extraction layer. The rent is no longer shared with the block builder and validator; it is captured by the solver and the aggregator. The user still pays, they just do not see the fee.
Contrarian: The real danger is not that intent-based systems replace DEXs—they will not, for the same reason dark pools never replaced open exchanges—but that they fragment the very concept of settlement finality. Today, a trade executed via an intent network is final only after the settlement transaction is included on-chain, but the user’s subjective experience of “the trade is done” happens when the solver confirms the intent. That gap, which can be anywhere from 1 to 30 seconds, is where a new class of settlement risk emerges. I have documented three cases in the last quarter where a solver accepted an intent, quoted a price based on a fleeting liquidity snapshot, and then the underlying pool moved sharply before the solver could hedge. The user received the quoted price—the solver absorbed the loss—but that loss is then socialised across future trades via wider spreads. The system is stable only as long as solvers have deep enough balance sheets to absorb shocks. In a black swan event, where multiple solvers face simultaneous adverse moves, the entire settlement layer could freeze. The network would revert to on-chain execution, but by then, trust in the off-chain promise would have evaporated.
Takeaway: The narrative that intent-based systems are the endgame for DeFi is itself a narrative trap. It offers the comfort of solved problems while hiding the emergence of new, less visible dependencies. Based on my experience tracking the evolution of exchange architectures since 2017, I believe the next inflection point will come not from technological refinement, but from a governance reckoning. Solver networks, aggregators, and relayers will face pressure to open their algorithms, publish their settlement data, and submit to periodic third-party audits. The industry will either codify a new standard of transparency for off-chain execution, or it will watch the same cycle repeat: early adopters profit from opacity, latecomers get burned, and regulators step in to impose a solution that no one wanted. Decoding the whisper before it becomes a shout requires us to ask not whether the system works today, but whether it will still hold when the storm comes. Navigating that storm with an anchor made of code means ensuring that every layer of trust is verifiable, not just optimised. Art is not just seen; it is verified and held. The same must be true for the infrastructure that moves our assets.
A quiet observation in a loud, decentralized room: the next war in crypto will not be about speed or fees. It will be about who gets to see the trade before it happens, and how we choose to define fairness when the solver knows more than the user. That is the narrative worth hunting.